By derivatizing the purely competitive CK2 inhibitor N1-(4,5,6,7-tetrabromo-1H-benzimidazol-2-yl)-propane-1,3-diamine (K137) at its 3-amino position with a peptidic fragment composed of three or four glutamic or aspartic acid residues, a new family of bisubstrate inhibitors has been generated whose ability to simultaneously interact with both the ATP and the phosphoacceptor substrate-binding sites has been probed by running mixed competition kinetics and by mutational mapping of the kinase residues implicated in substrate recognition. The most effective bisubstrate inhibitor, K137-E4, interacts with three functional regions of the kinase: the hydrophobic pocket close to the ATP-binding site, the basic residues of the p+1 loop that recognizes the acidic determinant at position n+1 and the basic residues of α-helixC that recognize the acidic determinant at position n+3. Compared with the parent inhibitor (K137), K137-E4 is severalfold more potent (IC50 25 compared with 130 nM) and more selective, failing to inhibit any other kinase as drastically as CK2 out of 140 enzymes, whereas 35 kinases are inhibited more potently than CK2 by K137. K137-E4 is unable to penetrate the cell and to inhibit endogenous CK2, its pro-apoptotic efficacy being negligible compared with cell-permeant inhibitors; however, it readily inhibits ecto-CK2 on the outer cell surface, reducing the phosphorylation of several external phosphoproteins. Inhibition of ecto-CK2 by K137-E4 is accompanied by a slower migration of cancer cells as judged by wound healing assays. On the basis of the cellular responses to K137-E4, we conclude that ecto-CK2 is implicated in cell motility, whereas its contribution to the pro-survival role of CK2 is negligible.

INTRODUCTION

Protein kinases are implicated in a wide variety of human diseases, with special reference to neoplasia. Consequently, compounds that target subsets of the >500 human protein kinases (commonly referred to as the human ‘kinome’) have become increasingly important in drug development. Most of these inhibitors are typically targeted at the ATP-binding site that is highly conserved across the members of the kinome. This hampers the development of highly selective inhibitors that are able to target individual kinases. A strategy to increase selectivity is based on the design of so-called ‘bisubstrate’ (or bifunctional) inhibitors consisting of two combined moieties, one targeted to the ATP-binding site, and the other shaped in such a way that it mimics the phosphoacceptor substrate and therefore competes for its binding site(s). At variance with ATP site-directed inhibitors, which sometimes bind with high affinity but often tend to be rather promiscuous, and with pseudosubstrate inhibitors, which in principle are quite selective but generally bind very weakly, bisubstrate inhibitors are expected to be both very potent and selective owing to co-operative effects of their dual mode of binding.

In the last decade, several successful examples of this strategy have been reported (reviewed in [1,2]), leading to the attainment of highly selective bisubstrate inhibitors with high affinity towards a variety of protein kinases, including PKA (protein kinase A), ROCK (Rho-associated kinase) and other basophilic kinases [37], the insulin receptor protein tyrosine kinase [8], PKC (protein kinase C) [9] and Akt (also known as protein kinase B) [10]. Bisubstrate inhibitors have also been developed to target protein kinase CK2, a very pleiotropic, constitutively active serine/threonine protein kinase [1113], whose abnormally high level in many tumours [14] has been shown to represent a remarkable example of non-oncogene addiction of cancer cells [15] and whose ATP site-directed inhibitor CX-4945 is presently in clinical trials for the treatment of a variety of malignancies [16]. Unlike the majority of serine/threonine protein kinases, CK2 is eminently acidophilic in nature, recognizing phosphoacceptor sites specified by multiple (on average more than five [12]) acidic side chains. By exploiting this unique property, the ATP site-directed inhibitor TBI (4,5,6,7-tetrabromo-1H-benzimidazole), whose structure in complex with CK2 had been previously solved [17], has been derivatized with polyaspartate fragments expected to mimic the phosphoacceptor substrate, thus giving rise to a remarkable increase in inhibitory potency compared with TBI [18].

Failure to determine the whole structure of the ligand appears to be a common feature of the crystal structures of complexes formed by bisubstrate inhibitors of CK2 as well as of other kinases. All attempts made in this respect have led to the detection of the ATP-competitive moiety alone, whereas the peptide moiety mimicking the phosphoacceptor substrate was not visible and not defined by electron density [7,18,19], arguing in favour of a flexible mode of binding whereby multiple interchangeable interactions co-operate to stabilize the complex. Therefore the pose of the whole bisubstrate inhibitor remained a matter of conjecture, based on the construction of suitable models and the assumption that the pose of the ATP-competitive moiety is not distorted by the interactions of the peptide segment.

We decided to adopt an alternative strategy based on the generation of CK2 mutants defective in their interactions with different elements of the bisubstrate inhibitor and to compare kinetically each of them with the wild-type for their susceptibility to inhibition. On the basis of these premises, we have exploited a TBI analogue, N′-(4,5,6,7-tetrabromo-1H-benzimidazol-2-yl)propane-1,3-diamine (K137) [20], a purely ATP competitive CK2 inhibitor, as the parent compound to generate a new class of bisubstrate inhibitors bearing segments of three or four glutamic or aspartic acid residues bound to its 3-amino position through a dicarboxyethyl linker. We show in the present paper that all of the derivatives tested (K137-E3, K137-E4 and K137-D4) inhibit CK2 holoenzyme and its isolated catalytic subunit much more potently than K137, by interacting not only with the ATP-binding site (where K137 also interacts), but also with basic residues which recognize the acidic determinants at positions n+1 and n+3 relative to the target residue of the peptide substrate. Such a network of multiple interactions not only increases the potency of the bisubstrate inhibitors, but also dramatically improves their selectivity as determined from profiling K137 and K137-E4 on a panel of 140 protein kinases. An added value of K137-E4 is its ability to inhibit only that pool of CK2 residing on the outer surface of the cell, thus providing a valuable tool to discriminate between ecto- and endo-cellular CK2.

EXPERIMENTAL

Materials

Recombinant CK2 and DYRK1A (dual-specificity tyrosine-phosphorylation-regulated kinase 1A) were purified as described in [21]. All mutants of the CK2α subunit were generated as reported in [22,23]. Recombinant CK1δ was purified as described in [24]. Recombinant GST–PLK1 (Polo-like kinase 1), GST–PLK2 and GST–PLK3 were expressed in Escherichia coli BL21 pLysS cells and purified as in [25]. PGEX6p1-Plk1 plasmid was described in [26]. Recombinant Fam20C and Erk8 were kindly provided by Dr Vincent Tagliabracci [27] and Dr Mario Chiariello [28] respectively. Recombinant GSK3β (glycogen synthase kinase 3β) was purchased from Sigma. The source of PIM-1, HIPK2 (homeodomain-interacting protein kinase 2) and all of the other protein kinases used for selectivity profiling is described in [29]. Dephosphorylated casein was purchased from Sigma. WGA (wheatgerm agglutinin) was from Life Technologies

Inhibitor 2 of protein phosphatase 1 was kindly provided by Professor Anna Depaoli-Roach. Synthetic peptide substrates were synthetized by the CRIBI (Centro di Ricerca Interdipartmentale per le Biotecnologie Innovative) Peptide Facility of the University of Padova. Anti-CK2 antibodies were raised in rabbits as reported in [22]. Procedure with animals were performed in accordance with the legal requirements of national authority. Anti-Hsp70 (heat-shock protein 70) antibodies were from Abcam. Secondary antibodies conjugated to horseradish peroxidase for Western blot development were purchased from PerkinElmer, whereas Alexa Fluor® 488 and 555-conjugated secondary antibodies were from Life Technologies. RPMI 1640 medium, DMEM (Dulbecco's modified Eagle's medium) and PBS were purchased from Sigma. Immobilon-P membranes were purchased from Millipore.

Chemistry

Peptidic derivatives were prepared according to a classical Fmoc solid-phase peptide synthetic strategy on Wang resin [30]. The synthetic path was designed in order to obtain the highest degree of purity. In particular, the peptidic fragments (Glu-Glu-Glu, Glu-Glu-Glu-Glu and Asp-Asp-Asp-Asp) were synthesized starting from the C-terminal amino acid linked to the resin. After the subsequent addition of the desired amino acids, the peptide obtained was coupled with N-[3-(4,5,6,7-tetrabromo-1H-benzoimidazol-2-ylamino)-propyl]-succinamic acid obtained from K137 [20].

As a final step, the product was cleaved from the resin and deprotected. The synthesized molecules were fully characterized by NMR and MS. Moreover, the purity profile was assessed by HPLC analysis.

Details concerning all compounds presented are provided in the Supplementary Online Data.

Molecular Modelling

The crystal structure of human CK2 was retrieved from the Protein Data Bank [31] (PDB code 3Q04) and processed in order to remove ligands and water molecules. Hydrogen atoms were added to the protein structure using standard geometries with the MOE (Molecular Operating Environment) program (www.chemcomp.com/MOE-Molecular_Operating_Environment.htm). To minimize contacts between hydrogens, the structures were subjected to Amber99 force-field minimization until the RMSD of the conjugate gradient was <0.1 kcal·mol−1·Å−1 (1 kcal=4.184 kJ; 1 Å=0.1 nm) keeping the heavy atoms fixed at their crystallographic positions. After the preparation phase, all compound structures were docked directly into the CK2 crystal structure, using the Glide package [32] of Schrödinger Suite 2012 (http://www.schrodinger.com/). The first stage of the in silico analysis was performed by running Site Map [33], implemented in the Schrödinger Suite; this tool highlights site points both in the hydrophobic ATP-binding zone as well as a basic solvent-exposed binding region matching with the substrate-interacting surface. Using the identified site points, a Glide grid (outer grid dimensions 15 Å×36 Å×20 Å) was built to perform the docking procedure. Ligand docking was carried out in the defined site by using the XP (extra precision) procedure. The best ranked ligand–protein complexes from docking experiments were subjected to refinement using the Prime MM-GBSA module [34] in Schrödinger Suite, considering only residue side chains 5 Å from the ligands.

In vitro kinase assays

Recombinant CK2 holoenzyme and mutant variants (1 nM) were incubated for 10 min at 37°C in a final volume of 25 μl containing 50 mM Tris/HCl (pH 7.5), 100 mM NaCl, 12 mM MgCl2, 100 μM synthetic peptide substrate CK2-tide (RRRADDSDDDDD), replaced in some experiments by 200 μM dephosphorylated casein or inhibitor 2 of protein phosphatase 1, and 20 μM [γ-33P]ATP (500–1000 c.p.m./pmol), unless indicated otherwise. The reaction was stopped by the addition of 5 μl of 0.5 M orthophosphoric acid before spotting 25 μl on to phosphocellulose filters. CK1δ (25 nM) was assayed as described in [35]. Activity of DYRK1A (23 nM) was measured according to [21]. PIM-1 (44 nM) was assayed as described in [36]. The HIPK2 (19 nM) phosphorylation assay was performed following the procedure described in [37]. PLK2 (100 nM) was assayed following the procedure described in [38,39]. PLK1 and PLK3 (100 nM) phosphorylation assays were performed following the same procedures used for PLK2 in the presence of 300 μM peptide substrate RRRISDELMDATFADQEAK. GSK3β was assayed as described in [40]. Activity of ERK8 (extracellular-signal-regulated kinase 8) (85 nM) and Fam20C (family with sequence similarity 20 member C) (25 nM) were measured according to [27] and [28] respectively. Conditions for the activity assays of all other protein kinases tested in selectivity experiments are as described or referenced in [29].

Kinetic determinations

Initial velocities were determined at each of the substrate concentration tested. Km and Vmax values were calculated either in the absence or in the presence of increasing concentrations of inhibitor, from Lineweaver–Burk double-reciprocal plots of the data. The inhibition constant for K137 was then calculated by linear regression analysis of Km/Vmax against inhibitor concentration plots. Inhibition constants for K137-E4 were calculated using eqn (3.2) for mixed-model inhibition in [41]

Selectivity profiles

Promiscuity scores (expressing the average inhibition of all of the kinases of the whole panel, by a given concentration of the inhibitor) and hit rates (expressing the percentage of kinases inhibited >50% by a given compound) were calculated from the selectivity data as described in [42,43].

Cell culture and treatments

Cells were cultured in an atmosphere containing 5% CO2; Jurkat cells were maintained in suspension in RPMI 1640 medium, HEK (human embryonic kidney)-293-T and HeLa cells were maintained in DMEM; both media were supplemented with 10% (v/v) FBS, 2 mM L-glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin. Cell treatments were performed in the culture medium with 1% FBS, at 106 cells/ml for cells that grow in suspension and ~70% confluence for adherent cells. Control cells were treated with equal amounts of the inhibitor solvent.

Cell lysis, Western blot experiments and measurement of CK2 activity in cells

Cells were harvested by centrifugation, washed and lysed as described in [44]. Protein concentration was determined using the Bradford method. Equal amounts of proteins were separated by SDS/PAGE (11% gels), blotted on Immobilon-P membranes, and processed by WB (Western blotting) with the indicated antibody, detected by chemiluminescence on a Kodak Image Station 440MM PRO. CK2 endocellular activity was evaluated by analysis of the CK2-dependent site Akt Ser129 phosphorylation [45]. Anti-total Akt was from Santa Cruz Biotechnology, anti-tubulin was from Sigma, anti-phospho-Ser129 Akt was raised in rabbits and purified as in [45].

Cell viability

For cell viability experiments, cell treatments were performed in 96-well plates at a density of 105 cells/100 μl for cells in suspension; 2×104 adherent cells were seeded in 96-well plates in 100 μl of medium, and left to adhere overnight before treatment. At 1 h before the end of the incubation, 10 μl of MTT solution (5 mg/ml in PBS) was added to each well. Incubations were stopped by addition of 20 μl of lysis solution at pH 4.7, containing 20% (w/v) SDS, 50% (v/v) N,N-dimethylformamide, 2% (v/v) acetic acid and 25 mM HCl. The A590 of plates was measured using a Titertek Multiskan Plus plate reader (Flow Laboratories).

CK2 ecto-kinase activity assay

For CK2 ecto-kinase activity assay, the method described in [46] was applied. After treatment, cells [(1–2)×105 HEK-293T or HeLa cells, (0.5–1)×105 Jurkat cells] were harvested, pelleted in sterile tubes, and washed three times with an isotonic buffer containing 70 mM NaCl, 30 mM Tris/HCl (pH 7.5), 5 mM MgCl2, 5 mM KH2PO4, 0.5 mM EDTA and 75 mM glucose. Then cells were resuspended in 50 μl of the same buffer and pre-incubated for 3 min at 37°C, before starting phosphorylation reactions by the addition of a mixture containing ATP (10 μM final concentration), [γ-33P]ATP (10000–20000 c.p.m./pmol) and CK2-tide (RRRADDSDDDDD peptide, 1 mM final concentration) [47], for a total volume of 56.6 μl. Samples incubated with the same mixture but without the peptide substrate were used to check non-specific phosphorylation. Controls of CK2-tide phosphorylation in the cell supernatant were performed under the same conditions using 50 μl of medium removed from centrifuged cells after the 3 min pre-incubation. Tubes were incubated at 37°C, with gentle shaking (130–150 rev./min) for various times. Then 20 μl of the phosphorylation mixture was spotted on to phosphocellulose paper and analysed as described in [47]. For the determination of the ecto-CK2/endo-CK2 ratio, CK2 activity of total Jurkat cell lysate was measured as in [48], normalizing the activity measured in 1–2 μg of cell proteins to the total proteins obtained from 106 cells.

Phosphorylation of ecto-CK2 substrates

HeLa cells (20000) were seeded in a 96-well plate and allowed to adhere. Cells were washed twice with an isotonic buffer containing 70 mM NaCl, 30 mM Tris/HCl (pH 7.5), 5 mM MgCl2, 5 mM KH2PO4, 0.5 mM EDTA and 75 mM glucose, then resuspended in 50 μl of the same buffer with or without K137-E4 or derivatives, and immediately incubated in a phosphorylation mixture containing ATP (10 μM final concentration), [γ-33P]ATP (10000–20000 c.p.m./pmol), and staurosporine (0.1 μM, in order to minimize unspecific phosphorylation) to a final volume of 56.6 μl. After 30 min of incubation in an atmosphere containing 5% CO2 at 37°C, 40 μl of the supernatant of each well was collected. Cells inside the plate were lysed with 10 μl of an ice-cold buffer consisting of 20 mM Tris/HCl (pH 7.5), 150 mM NaCl, 2 mM EDTA, 2 mM EGTA, 1% (v/v) Triton X-100, 2 mM DTT, Complete™ protease inhibitor cocktail (Roche), 10 mM NaF, 1 μM okadaic acid, 1 mM sodium vanadate and 4 mM ATP to block any phosphorylation reactions. After 1 h of shaking on ice, the lysates were collected from each well and centrifuged at 14220 g for 10 min, at 4°C. Proteins of the supernatants and of total cell lysates were analysed by SDS/PAGE and WB and digital autoradiography (CyclonePlus Storage Phosphor System, PerkinElmer).

CK2 immunofluorescence staining

HeLa cells (10000) were seeded on glass coverslips and allowed to adhere overnight. For ecto-CK2 staining, cells were washed gently once with PBS and incubated with the primary anti-CK2 antibody diluted at 1:50 in 2% (v/v) goat serum and 0.5% BSA/PBS solution for 2 h at room temperature. Then cells were washed once with PBS and incubated with Alexa Fluor® 488-conjugated goat anti-rabbit IgG secondary antibody at 1:500 dilution in the same solution used for primary antibody, for 1 h at room temperature and fixed with 4% (w/v) paraformaldehyde for 20 min at room temperature in the dark. As a control for ecto-CK2 localization, HeLa cell membrane staining was performed by means of WGA, a receptor ligand used for general labelling of plasma membranes. Cells were incubated for 1 h with the Alexa Fluor® 555-conjugated WGA at 1:50 dilution, together with the fluorescent secondary antibody, in the dark. For endo-CK2 staining, cells were fixed with 4% (w/v) paraformaldehyde for 20 min at room temperature, permeabilized with 0.1% Triton X-100 in PBS for 10 min at 4°C, and incubated overnight at 4°C with rabbit anti-CK2 antibody together with mouse anti-Hsp70 antibody, at 1:50 overnight. After three washes with PBS, cells were incubated with the fluorescent secondary antibodies Alexa Fluor® 555-conjugated goat anti-rabbit IgG and Alexa Fluor® 488-conjugated goat anti-mouse IgG, at 1:500 for 1 h. The coverslips were mounted with Mowiol (Sigma) solution and the immunofluorescence images were acquired with a Leica TCS SP5 confocal microscope equipped with HCX PL APO ×100.0 1.4 numerical aperture oil-immersion objective. LAS IF software was used for image processing.

Wound-healing assay

Cell migration was evaluated using an in vitro wound-healing assay. Identical wound areas were created into the cell monolayer by using Ibidi culture-inserts, according to the manufacturer's instructions. Briefly, 3×104 HeLa cells were seeded in complete DMEM on each side of the Ibidi culture-insert in a 24-well plate. After cell attachment, the medium was gently removed and replaced by 1% (v/v) FBS-supplemented DMEM containing increasing concentrations of K137-E4, or AS-E4, K137-E4Me or vehicle, as negative controls. After 16 h, the culture insert was detached in order to form a cell-free gap into the cell monolayer. Each well was rinsed once with PBS to remove cell debris, and immediately refilled with 1% (v/v) FBS-supplemented DMEM, containing the inhibitor or the vehicle. Cells were allowed to migrate for 24 h. The wound images were captured at the moment of the insert removal (t=0 h) and 24 h later (t=24 h) using a Leica DMI4000 automated inverted microscope equipped with a Leica DFC300 FX camera. The wound area covered by cells was calculated using Image J 1.47c software (http://rsb.info.nih.gov).

RESULTS

Mapping interactions with CK2

In order to generate compounds able to inhibit CK2 by simultaneously interacting with both its ATP- and phosphoacceptor substrate-binding site, the ATP site-directed inhibitor K137 has been derivatized by linking to its amino group a peptide chain composed of three or four aspartic or glutamic acid residues (see the Experimental section).

The three compounds obtained, denoted K137-E3, K134-E4 and K134-D4, were compared with the parent compound K137 for their ability to inhibit CK2 in a dose-dependent manner. The resulting IC50 values show that all three compounds are able to inhibit the CK2 holoenzyme with an efficacy higher than that of K137, the most effective inhibitor being the tetraglutamyl one (IC50=25 nM), followed by the tetra-aspartyl and the triglutamyl ones respectively (Table 1). By replacing the phosphoacceptor peptide with protein substrates, notably casein and inhibitor 1 of protein phosphatase 2, similar IC50 values were obtained, as mentioned in the legend of Table 1. The potency of the inhibitors is comparable with that of a series of similar compounds obtained by conjugating TBI with peptides composed of five or six aspartic acid residues (ARC-1154 and ARC-1183), exhibiting IC50 values of 83 and 30 nM respectively [18]. It is worth noting that the oligo-glutamyl and -aspartyl moieties alone are devoid of efficacy up to 40 μM concentration, highlighting a synergistic effect that depends on their covalent association with the ‘anchoring’ element, K137. On the other hand, the crucial role of their negative charges is documented by the loss of inhibitory efficacy of the tetramethyl ester of K137-E4 (K137-E4Me).

Table 1
IC50 values (μM) of selected compounds against CK2 holoenzyme

Results are means of experiments run in triplicate, with the S.E.M. never exceeding 10%. Synthesis of compounds is described in the Supplementary Online Data. CK2 activity was assayed using the specific peptide as phosphoacceptor substrate, as described in the Experimental section. Similar IC50 values were obtained by testing the inhibitor K137-E4 in a reaction medium where the peptide was replaced by either casein (IC50=0.031 μM) or inhibitor 2 of protein phosphatase 1 (IC50=0.026 μM).

Compound Formula IC50 CK2α2β2 Compound Formula IC50 CK2α2β2 
K137  0.13                        
K137-E3  0.033 AS-E3  >40 
K137-D4  0.035 AS-D4  >40 
K137-E4  0.025 AS-E4  >40 
K137-E4Me  10.0                                   
Compound Formula IC50 CK2α2β2 Compound Formula IC50 CK2α2β2 
K137  0.13                        
K137-E3  0.033 AS-E3  >40 
K137-D4  0.035 AS-D4  >40 
K137-E4  0.025 AS-E4  >40 
K137-E4Me  10.0                                   

Whereas the mode of inhibition of K137, similar to all the tetrabromo-benzimidazole analogues tested to date, is purely competitive with respect to ATP and non-competitive with respect to the phosphoacceptor peptide substrate (Figures 1A and 1B), the kinetics with K137-E4 and K137-D4 denote a mixed mode of inhibition with both the Vmax and the Km values (either for ATP or the peptide substrate) altered by increasing the inhibitor concentration (Figures 1C and 1D). It is clear from these kinetics that K137-E4 competes against both ATP and the peptide substrate binding, the former effect being more pronounced than the latter.

Inhibition of CK2 holoenzyme by K137 (A and B) and K137-E4 (C and D)

Figure 1
Inhibition of CK2 holoenzyme by K137 (A and B) and K137-E4 (C and D)

Kinetics were performed with respect to ATP or the peptide substrate CK2-tide in the absence or in the presence of the indicated inhibitor concentrations. In the case of K137-E4 mixed competitive kinetics, an inset showing the Vmax and Km variations at increasing inhibitor concentration is also included

Figure 1
Inhibition of CK2 holoenzyme by K137 (A and B) and K137-E4 (C and D)

Kinetics were performed with respect to ATP or the peptide substrate CK2-tide in the absence or in the presence of the indicated inhibitor concentrations. In the case of K137-E4 mixed competitive kinetics, an inset showing the Vmax and Km variations at increasing inhibitor concentration is also included

This result by itself supports the dual mode of binding of the new compounds. A more detailed analysis of their multiple interactions was performed by exploiting a number of CK2 mutants defective either in ability to bind ATP site-directed inhibitors or in substrate recognition.

As shown in Table 2, inhibition by K137 is compromised by mutating Val66 and Ile174, two unique CK2 residues which are close to the ATP-binding site and were shown to be essential for the high-affinity binding of many ATP site-directed CK2 inhibitors, with special reference to those which, similarly to K137, display a tetrabromo-benzimidazole scaffold [42]. Conversely, the mutation of residues which were shown to interact with the crucial acidic determinants at positions n+3 (Lys74–Lys77) and n+1 (Arg191, Arg195 and Lys198) in the phosphoacceptor substrate [22] has no effect on the inhibition by K137. In contrast, inhibition by K137-E3, K137-D4 and K137-E4 is invariably reduced by all of these mutations, either affecting the hydrophobic region close to the ATP site or the basic residues responsible for substrate recognition. Interestingly, the contribution of the hydrophobic residues, which is predominant and comparable with that observed with K137 in the case of the triglutamic derivative, becomes much less evident with the tetraglutamic one (88–17-fold fall in efficacy) with a concomitantly increased relevance of the basic residues interacting with the acidic determinant at the n+3 position in the phosphoacceptor substrate (the Lys74–Lys77 cluster). On the other hand the basic elements of the p+1 loop, responsible for the binding of the acidic determinant at n+1, are equally important with all the new compounds, regardless of the number and nature of the acidic residues forming their tail. These data suggest that, whereas all three new compounds interact with the p+1 loop, only K137-E4 and, to a lesser extent, its 4D homologue, can make high-affinity contacts with the basic cluster at the beginning of α-helixC.

Table 2
Mutational analysis of residues responsible for susceptibility of CK2 to K137 and derived bisubstrate inhibitors

Assays were performed as described in the Experimental section in the presence of 400 μM peptide substrate CK2-tide. IC50 values (μM) were determined using either CK2 wild-type or mutants altered in the hydrophobic pocket close to the ATP-binding site (V66A/I174A) [23] or in regions interacting with the crucial specificity determinants at position n+1 (R191A/R195A/K198A) and n+3 (K74A/K75A/K76A/K77A) respectively [22]. Fold variations of IC50 values of the mutants with respect to wild-type are indicated in parentheses.

Compound CK2α2β2 CK2α2(I174A/V66A)β2 CK2α2(K74A/K75A/K76A/K77A)β2 CK2α2(R191A/R195A/K198A)β2 
K137 0.16 19.1 (×117) 0.16 (×1) 0.17 (×1) 
K137-E3 0.038 3.39 (×88) 0.23 (×5.9) 0.34 (×8.8) 
K137-D4 0.033 1.57 (×47) 0.28 (×8.4) 0.30 (×9.1) 
K137-E4 0.026 0.460 (×17.1) 0.37 (×13.5) 0.22 (×8.3) 
Compound CK2α2β2 CK2α2(I174A/V66A)β2 CK2α2(K74A/K75A/K76A/K77A)β2 CK2α2(R191A/R195A/K198A)β2 
K137 0.16 19.1 (×117) 0.16 (×1) 0.17 (×1) 
K137-E3 0.038 3.39 (×88) 0.23 (×5.9) 0.34 (×8.8) 
K137-D4 0.033 1.57 (×47) 0.28 (×8.4) 0.30 (×9.1) 
K137-E4 0.026 0.460 (×17.1) 0.37 (×13.5) 0.22 (×8.3) 

In any case, it is clear from the data in Table 2 that both pseudosubstrate-like and ATP site-directed interactions contribute to the efficacy of the newly developed inhibitors and that their relative relevance varies in opposite ways. This is the first case, to the best of our knowledge, where the dual mode of binding of bisubstrate inhibitors has been unambiguously mapped, given the failure to detect pseudosubstrate interactions in the crystal structure of kinase–bisubstrate inhibitor complexes solved to date.

Selectivity

To evaluate the selectivity of the most effective bisubstrate inhibitor, K137-E4 has been profiled at 1 μM concentration on a panel of 140 protein kinases. As shown in Table 3, only CK2 activity is entirely suppressed by K137-E4 (residual activity <5%), with only one other kinase, ERK8, inhibited >70% (residual activity 25%). From these data, quite low values of promiscuity score (12.02) and hit rate (0.05) can be drawn, both denoting very high selectivity. Such a narrow specificity is clearly conferred by the tetraglutamic moiety of K137-E4, since the selectivity profile of K137, also reported in Table 3, reveals a remarkable promiscuity, with 37 kinases inhibited >50%, reflected in higher promiscuity scores (32.12) and hit rates (0.27). The higher selectivity of K137-E4 with respect to K137 is highlighted in a pictorial manner in Figure 2(A), where their selectivity maps are compared. In Figure 2(B), the residual activities with kinases inhibited >50% by K137 and/or K137-E4 are combined and presented as an individual histogram. It should be noted that almost all kinases which are >50% inhibited by K137 are either unaffected or much less inhibited by K137-E4. In contrast, only two kinases, CK2 and PLK1 are inhibited more effectively by K137-E4 than they are by K137; both are acidophilic in nature.

Comparison between K137 and K137-E4 selectivity

Figure 2
Comparison between K137 and K137-E4 selectivity

(A) Inhibition maps of K137 (left) and K137-E4 (right). (B) K137 (white) and K137-E4 (black) inhibition profiles of kinases with residual activity ≤50%. Values plotted are from Table 3.

Figure 2
Comparison between K137 and K137-E4 selectivity

(A) Inhibition maps of K137 (left) and K137-E4 (right). (B) K137 (white) and K137-E4 (black) inhibition profiles of kinases with residual activity ≤50%. Values plotted are from Table 3.

Table 3
Selectivity profiles of K137-E4 and K137 on a 140 kinase panel

Residual CK2 activity (determined at 1 μM inhibitor concentration) is expressed as a percentage of the control activity without inhibitor. Conditions are described or referenced in the Experimental section. Activities <50% of control are in bold.

 Activity (%) 
Kinase K137-E4 K137 
CK2 4 45 
ERK8 25 3 
DYRK2 40 2 
ULK1 41 22 
PLK1 44 85 
JAK2 46 62 
CLK2 48 3 
VEGFR 51 21 
GSK3β 57 98 
DYRK1A 57 7 
HIPK2 60 6 
DYRK3 60 7 
EIF2AK3 65 64 
TTK 67 58 
GCK 70 56 
TAK1 72 50 
PAK4 73 88 
TIE2 75 97 
MAPKAPK3 75 102 
SYK 77 82 
ZAP70 78 98 
PDGFRα 78 109 
ROCK 2 78 49 
CK1δ 78 77 
TTBK1 78 98 
CDK2–cyclin A 79 41 
MKK1 79 48 
IRR 79 23 
LKB1 81 93 
CDK9–cyclin T1 81 12 
Src 81 112 
PIM-3 82 21 
PAK2 83 100 
Lck 83 107 
MARK4 83 89 
BTK 84 77 
PRAK 84 92 
PKBβ 85 80 
PKD1 85 8 
RSK2 85 81 
TrkA 85 53 
DAPK1 85 47 
MKK6 85 88 
ULK2 86 55 
AMPK (hum) 86 102 
Aurora B 87 56 
SmMLCK 88 75 
TGFβR1 88 93 
PINK 88 117 
MELK 88 60 
PAK6 88 106 
CHK2 88 50 
TBK1 88 104 
FGFR1 89 91 
ERK2 89 78 
CAMK1 89 64 
TESK1 89 101 
SIK2 90 96 
EF2K 90 101 
STK33 90 38 
ABL 90 111 
SGK1 91 62 
p38α MAPK 91 98 
MLK3 91 29 
PIM-1 91 1 
MLK1 91 46 
EPH-B2 91 88 
RIPK2 92 86 
MSK1 92 48 
NEK6 92 59 
SIK3 92 107 
BRSK1 92 33 
CHK1 92 79 
MEKK1 93 119 
DDR2 93 105 
MARK2 93 75 
MNK2 93 44 
PRK2 93 32 
EPH-A2 94 86 
CK1γ2 94 94 
MKK2 94 73 
S6K1 94 29 
MARK1 95 64 
IRAK1 95 23 
EPH-B1 95 83 
RSK1 95 80 
CAMKKβ 96 97 
PIM-2 96 20 
JNK2 96 90 
HIPK3 96 11 
SRPK1 96 105 
ASK1 96 97 
TTBK2 97 105 
YES1 97 106 
MAP4K3 97 107 
WNK1 97 104 
IKKε 97 86 
PAK5 97 111 
PKBα 97 11 
NUAK1 97 51 
CSK 97 88 
IRAK4 98 60 
MARK3 98 80 
IR 98 27 
OSR1 98 113 
BRK 99 105 
TSSK1 99 92 
Aurora A 99 92 
IKKβ 99 46 
MAPKAPK2 99 93 
HIPK1 99 7 
TLK1 100 99 
MPSK1 100 97 
NEK2a 101 92 
TAO1 101 66 
MINK1 101 79 
MNK1 102 51 
p38δ MAPK 102 88 
EPH-A4 102 84 
PDK1 102 92 
PKCζ 102 13 
MST3 103 97 
IGF-1R 103 27 
PHK 103 36 
MST4 103 75 
HER4 104 92 
ERK5 104 38 
JNK3 104 76 
PKCγ 104 105 
JNK1 105 77 
MST2 105 78 
EPH-B4 106 107 
p38γ MAPK 107 79 
ERK1 107 96 
MAP4K5 109 81 
PKCα 110 94 
PKA 112 68 
p38β MAPK 114 87 
EPH-B3 118 68 
BRSK2 132 39 
 Activity (%) 
Kinase K137-E4 K137 
CK2 4 45 
ERK8 25 3 
DYRK2 40 2 
ULK1 41 22 
PLK1 44 85 
JAK2 46 62 
CLK2 48 3 
VEGFR 51 21 
GSK3β 57 98 
DYRK1A 57 7 
HIPK2 60 6 
DYRK3 60 7 
EIF2AK3 65 64 
TTK 67 58 
GCK 70 56 
TAK1 72 50 
PAK4 73 88 
TIE2 75 97 
MAPKAPK3 75 102 
SYK 77 82 
ZAP70 78 98 
PDGFRα 78 109 
ROCK 2 78 49 
CK1δ 78 77 
TTBK1 78 98 
CDK2–cyclin A 79 41 
MKK1 79 48 
IRR 79 23 
LKB1 81 93 
CDK9–cyclin T1 81 12 
Src 81 112 
PIM-3 82 21 
PAK2 83 100 
Lck 83 107 
MARK4 83 89 
BTK 84 77 
PRAK 84 92 
PKBβ 85 80 
PKD1 85 8 
RSK2 85 81 
TrkA 85 53 
DAPK1 85 47 
MKK6 85 88 
ULK2 86 55 
AMPK (hum) 86 102 
Aurora B 87 56 
SmMLCK 88 75 
TGFβR1 88 93 
PINK 88 117 
MELK 88 60 
PAK6 88 106 
CHK2 88 50 
TBK1 88 104 
FGFR1 89 91 
ERK2 89 78 
CAMK1 89 64 
TESK1 89 101 
SIK2 90 96 
EF2K 90 101 
STK33 90 38 
ABL 90 111 
SGK1 91 62 
p38α MAPK 91 98 
MLK3 91 29 
PIM-1 91 1 
MLK1 91 46 
EPH-B2 91 88 
RIPK2 92 86 
MSK1 92 48 
NEK6 92 59 
SIK3 92 107 
BRSK1 92 33 
CHK1 92 79 
MEKK1 93 119 
DDR2 93 105 
MARK2 93 75 
MNK2 93 44 
PRK2 93 32 
EPH-A2 94 86 
CK1γ2 94 94 
MKK2 94 73 
S6K1 94 29 
MARK1 95 64 
IRAK1 95 23 
EPH-B1 95 83 
RSK1 95 80 
CAMKKβ 96 97 
PIM-2 96 20 
JNK2 96 90 
HIPK3 96 11 
SRPK1 96 105 
ASK1 96 97 
TTBK2 97 105 
YES1 97 106 
MAP4K3 97 107 
WNK1 97 104 
IKKε 97 86 
PAK5 97 111 
PKBα 97 11 
NUAK1 97 51 
CSK 97 88 
IRAK4 98 60 
MARK3 98 80 
IR 98 27 
OSR1 98 113 
BRK 99 105 
TSSK1 99 92 
Aurora A 99 92 
IKKβ 99 46 
MAPKAPK2 99 93 
HIPK1 99 7 
TLK1 100 99 
MPSK1 100 97 
NEK2a 101 92 
TAO1 101 66 
MINK1 101 79 
MNK1 102 51 
p38δ MAPK 102 88 
EPH-A4 102 84 
PDK1 102 92 
PKCζ 102 13 
MST3 103 97 
IGF-1R 103 27 
PHK 103 36 
MST4 103 75 
HER4 104 92 
ERK5 104 38 
JNK3 104 76 
PKCγ 104 105 
JNK1 105 77 
MST2 105 78 
EPH-B4 106 107 
p38γ MAPK 107 79 
ERK1 107 96 
MAP4K5 109 81 
PKCα 110 94 
PKA 112 68 
p38β MAPK 114 87 
EPH-B3 118 68 
BRSK2 132 39 

To obtain a deeper insight into this aspect, the dose-dependent inhibition by K137-E4 and K137 of these and other acidophilic kinases was determined. In this assay, a few acidophilic kinases were also included which were not present in the panel of Table 4, notably PLK2 and PLK3 and the atypical kinase Fam20C which has been recently shown to account for the catalytic activity of GCK (genuine casein kinase), responsible for the phosphorylation of casein as well as of a plethora of other secreted proteins at S-X-E motifs [27,49,50]. The IC50 values for the inhibition of these and other representative kinases by either K137 or K137-E4 are reported in Table 4. These values confirm the data presented in Figure 2(B) as far as CK2 and PLK1, on one side, and ERK, PIM-1, HIPK2 and DYRK1A, on the other, are concerned; they also show that higher susceptibility to K137-E4 than to K137 is shared by a few other acidophilic kinases (not included in the panel of Table 3), with special reference to PLK2 and PLK3.

Table 4
IC50 values (μM) of K137 and K137-E4 calculated with different protein kinases

Results are means for experiments run in triplicate, with the S.E.M. never exceeding 10%. The ratio between the two values is shown in the bottom line.

Compound CK2α2β2 CK2α PLK1 PLK2 PLK3 ERK8 GSK3β Fam20C CK1δ HIPK2 PIM-1 DYRK1A 
K137 0.13 0.25 >40 39.0 27.0 0.212 0.752 >40 5.7 0.34 0.25 0.23 
K137-E4 0.025 0.030 0.420 0.400 0.532 0.854 1.554 >40 16 >40 >40 1.26 
K137/K137-E4 5.2 8.3 >95.0 97.5 50.7 0.24 0.48 – 0.35 <0.008 <0.006 0.18 
Compound CK2α2β2 CK2α PLK1 PLK2 PLK3 ERK8 GSK3β Fam20C CK1δ HIPK2 PIM-1 DYRK1A 
K137 0.13 0.25 >40 39.0 27.0 0.212 0.752 >40 5.7 0.34 0.25 0.23 
K137-E4 0.025 0.030 0.420 0.400 0.532 0.854 1.554 >40 16 >40 >40 1.26 
K137/K137-E4 5.2 8.3 >95.0 97.5 50.7 0.24 0.48 – 0.35 <0.008 <0.006 0.18 

These data corroborate the view that PLKs are also acidophilic kinases susceptible therefore to a bisubstrate mode of inhibition by K137-E4, although in this case the affinity for the hydrophobic (ATP-competitive) moiety (K137) is much weaker than in the case of CK2. Indeed, at variance with CK2, PLKs display a very modest sensitivity to K137, suggesting that their inhibition by K137-E4 is mainly due to the tetraglutamic (‘E4’) moiety of the compound. Consistent with this interpretation kinetics of inhibition of PLK2 by K137-E4 are also of mixed type with respect to both ATP and the peptide substrate, revealing, however, a competition by the peptide substrate more pronounced than by ATP (results not shown).

Modelling

Molecular docking experiments of K137-E4 confirm its double mode of binding by which the hydrophobic part of the molecule, represented by the K137 moiety, is bound to the ATP-binding cleft, whereas the acidic segment stretches out to the substrate-binding zone (Figure 3). In particular, the hydrophobic moiety establishes interactions with Val66, Val53, Ile174 and Asp175, whereas the four glutamic acid residues interact with basic amino acids whose role in substrate recognition is well known [22]. The first glutamic acid residue interacts with Lys77 located in α-helixC, and to a lesser extent, with Lys49 of the p-loop; the second and third interact with the p+1 loop in particular with Lys198 and Arg195 respectively; and the fourth one interacts with Lys74, another important basic element of α-helixC, and Arg191. All of the other bisubstrate derivatives establish similar interactions (results not shown); however, K137-E4 is the only compound able to interact with both of the basic residues of α-helixC (Lys74 and Lys77), recognizing the n+3 position of the phosphoacceptor site. The resulting model is in good agreement with the experimental data obtained in vitro using recombinant mutants (Table 2), showing that K137-E4 is able to form at least one more effective interaction with CK2 substrate-binding zone (Lys74), compared with the other bisubstrate inhibitors. Consequently, this compound is also the one least sensitive to the double mutation V66A/I174A, while being the most sensitive to the K74A/K75A/K76A/K77A mutation (Table 2), as expected assuming a stronger interaction of K137-E4 with the phosphoacceptor substrate-binding zone compared with the other derivatives.

In silico interaction between K137-E4 and CK2

Figure 3
In silico interaction between K137-E4 and CK2

K137-E4 (yellow) was docked directly into CK2 crystal structure (PDB code 3Q04) as described in the Experimental section; the residues directly involved in the interaction are highlighted, and the IC50 values of corresponding alanine mutants are indicated for comparison with that of wild-type CK2 (0.026 μM, see also Table 2). K74-77A, K74A/K75A/K76A/K77A.

Figure 3
In silico interaction between K137-E4 and CK2

K137-E4 (yellow) was docked directly into CK2 crystal structure (PDB code 3Q04) as described in the Experimental section; the residues directly involved in the interaction are highlighted, and the IC50 values of corresponding alanine mutants are indicated for comparison with that of wild-type CK2 (0.026 μM, see also Table 2). K74-77A, K74A/K75A/K76A/K77A.

Inhibition of ecto-CK2 by cell treatment with K137-E4

Upon treatment of Jurkat cells with 10 μM K137, endogenous CK2 activity is strongly inhibited, as determined from the decrease in Akt Ser129 phosphorylation, similar to that observed upon treatment with CX-4945, the first-in-class CK2 inhibitor available to date (Figure 4). By sharp contrast, as also shown in Figure 4, cell treatment with K137-E4 has no effect on endogenous CK2 activity, consistent with the concept that its acidic tail prevents permeability through the cell membrane. This observation prompted us to check whether ecto-CK2, i.e. that pool of CK2 which is externalized on the outside of many cell membranes [5153] might be selectively inhibited by K137-E4. To this purpose, the ecto-kinase CK2 activity of Jurkat and HEK-293T cells was monitored as described in the Experimental section. As shown in Figure 5(A), this method allows us to measure the CK2 activity present on the outer side of washed cells suspended and incubated with [γ-33P]ATP in the presence of a specific CK2 peptide substrate unable to penetrate the cell. Under these conditions, the activity of CK2 in the cell supernatant is negligible compared with the whole ‘external’ CK2 activity, thus providing the evidence that the majority of this is a bona fide ecto-kinase activity, owing to CK2 bound to the outer surface of the cell and not to CK2 released into the medium due to accidental cell lysis occurring during manipulation. At variance with endogenous CK2, which is inhibited upon cell treatment with K137 and CX-4945 but not with K137-E4 (Figure 4), this ecto-CK2 activity is suppressed by K137-E4 (Figure 5B) as effectively as by K137 and CX-4945 (results not shown). Similar results were also obtained with HeLa cells (not shown). It can be concluded therefore that, although K137-E4 is not cell-permeant, it represents a valuable tool for specifically targeting that pool of CK2 which is present on the cell surface, without affecting intracellular CK2. To gain information about the relevance of this subset of externalized CK2 molecules, we made a quantitative estimate of the relative amounts of ecto- and endo-CK2 in two different cell types. Figure 6(A) shows that ecto-CK2 activity in Jurkat cells is approximately one-tenth of the total CK2 activity measurable in cell lysates. By applying an immunofluorescence approach, where ecto-CK2 is visualized by the anti-CK2 antibody reactivity in intact unpermeabilized cells, it was also possible to show the existence of a pool of extracellular membrane-bound CK2 in HeLa cells (Figure 6B).

CK2 activity in Jurkat cells treated with CK2 inhibitors

Figure 4
CK2 activity in Jurkat cells treated with CK2 inhibitors

Total proteins (20 μg) from Jurkat cells treated for 24 h as indicated was analysed by WB with an antibody towards phospho-Ser129 Akt, a CK2-specific site [47]; total Akt and actin WB, analysed for normalization, are also shown. Western blots are representative of five independent experiments. The histograms represent the normalized quantification of the phospho-Ser129 bands, where a value of 100 has been assigned to the signal of untreated cells.

Figure 4
CK2 activity in Jurkat cells treated with CK2 inhibitors

Total proteins (20 μg) from Jurkat cells treated for 24 h as indicated was analysed by WB with an antibody towards phospho-Ser129 Akt, a CK2-specific site [47]; total Akt and actin WB, analysed for normalization, are also shown. Western blots are representative of five independent experiments. The histograms represent the normalized quantification of the phospho-Ser129 bands, where a value of 100 has been assigned to the signal of untreated cells.

CK2 ecto-kinase activity is abrogated by K137-E4

Figure 5
CK2 ecto-kinase activity is abrogated by K137-E4

(A) CK2 activity was measured in the presence (+) or the absence (−) of the synthetic peptide CK2-tide for the indicated times. Intact cells or supernatant after cell centrifugation were assayed as indicated (see the Experimental section for details). Results are means±S.E.M., expressed as c.p.m., of four independent experiments. (B) CK2 ecto-kinase activity of the indicated intact cells towards CK2-tide was measured for 30 min and subtracted from the activity obtained in the absence of CK2-tide. K137-E4 was added, as indicated, during the phosphorylation assay. Activity is reported as the percentage of the control, obtained in the absence of K137-E4.

Figure 5
CK2 ecto-kinase activity is abrogated by K137-E4

(A) CK2 activity was measured in the presence (+) or the absence (−) of the synthetic peptide CK2-tide for the indicated times. Intact cells or supernatant after cell centrifugation were assayed as indicated (see the Experimental section for details). Results are means±S.E.M., expressed as c.p.m., of four independent experiments. (B) CK2 ecto-kinase activity of the indicated intact cells towards CK2-tide was measured for 30 min and subtracted from the activity obtained in the absence of CK2-tide. K137-E4 was added, as indicated, during the phosphorylation assay. Activity is reported as the percentage of the control, obtained in the absence of K137-E4.

Quantification of ecto- and endo-CK2

Figure 6
Quantification of ecto- and endo-CK2

(A) CK2 activity was measured towards the synthetic peptide CK2-tide in intact Jurkat cells [(0.5–1)×105 cells, ecto-CK2] and in total lysate from Jurkat cells (1–2 μg, endo-CK2) (see the Experimental section for details). The activity was linear with the amount of cells or lysate used, and is shown after normalization to 106 cells. (B) Upper panels (ecto-CK2): immunofluorescence CK2 staining was performed in HeLa cells omitting any cell permeabilization procedure, to permit the anti-CK2 antibody to react only at the external cellular side; the secondary antibody used was Alexa Fluor® 488-conjugated anti-rabbit IgG (green), whereas WGA staining (red) is shown as marker of plasma membrane. Lower panels (endo-CK2): HeLa cells were permeabilized and incubated with anti-CK2 (Alexa Fluor® 555-conjugated anti-rabbit IgG secondary antibody, red), whereas anti-Hsp70 (Alexa Fluor® 488-conjugated anti-mouse IgG secondary antibody, green), was used as marker of an endocellular protein. In merged images, yellow signal denotes co-localization. All secondary antibodies used gave negligible signals when incubated in the absence of primary antibodies (results not shown).

Figure 6
Quantification of ecto- and endo-CK2

(A) CK2 activity was measured towards the synthetic peptide CK2-tide in intact Jurkat cells [(0.5–1)×105 cells, ecto-CK2] and in total lysate from Jurkat cells (1–2 μg, endo-CK2) (see the Experimental section for details). The activity was linear with the amount of cells or lysate used, and is shown after normalization to 106 cells. (B) Upper panels (ecto-CK2): immunofluorescence CK2 staining was performed in HeLa cells omitting any cell permeabilization procedure, to permit the anti-CK2 antibody to react only at the external cellular side; the secondary antibody used was Alexa Fluor® 488-conjugated anti-rabbit IgG (green), whereas WGA staining (red) is shown as marker of plasma membrane. Lower panels (endo-CK2): HeLa cells were permeabilized and incubated with anti-CK2 (Alexa Fluor® 555-conjugated anti-rabbit IgG secondary antibody, red), whereas anti-Hsp70 (Alexa Fluor® 488-conjugated anti-mouse IgG secondary antibody, green), was used as marker of an endocellular protein. In merged images, yellow signal denotes co-localization. All secondary antibodies used gave negligible signals when incubated in the absence of primary antibodies (results not shown).

Few ecto-CK2 substrates have been identified so far [5457]; however, as shown in Figure 7, several external phosphoproteins seem to be dependent on its activity, since their radiolabelling becomes evident upon cell incubation with [γ-33P]ATP and is reduced by the addition of K137-E4. Note that neither the tetraglutamyl moiety of K137 (AS-E4) nor the tetramethyl ester of K137-E4 (K137-E4Me) are able to reduce the radioactive phosphorylation of these proteins, consistent with the in vitro data shown in Table 1, and suggesting that ecto-CK2 is majorly responsible for their phosphorylation, under our experimental conditions. We found very similar results by loading the cellular pellets after intact cell phosphorylation, in order to analyse the labelling of ecto-substrates still bound to the cell surface (not shown). These experiments disclose the potential of our compound for future identification of new ecto-CK2 substrates.

Effects of K137-E4 on cellular targets of ecto-CK2

Figure 7
Effects of K137-E4 on cellular targets of ecto-CK2

Intact HeLa cells were incubated with a phosphorylation mixture in an isotonic buffer, in the absence or presence of K137-E4, AS-E4 or K137-E4Me, as indicated. After 30 min of incubation, aliquots of cellular medium were collected and separated by SDS/PAGE, blotted and analysed by digital autoradiography. Proteins (10 μg) from cellular lysates were used for normalization by WB with anti-actin antibody.

Figure 7
Effects of K137-E4 on cellular targets of ecto-CK2

Intact HeLa cells were incubated with a phosphorylation mixture in an isotonic buffer, in the absence or presence of K137-E4, AS-E4 or K137-E4Me, as indicated. After 30 min of incubation, aliquots of cellular medium were collected and separated by SDS/PAGE, blotted and analysed by digital autoradiography. Proteins (10 μg) from cellular lysates were used for normalization by WB with anti-actin antibody.

Although the role of this externalized CK2 pool is still unclear (as well as that of ecto-kinase activities in general), the experiment reported in Figure 8 rules out its implication in the well-known global anti-apoptotic function of CK2, by showing that, whereas cell treatment with increasing concentrations of CX-4945 and K137 (both cell-permeant CK2 inhibitors) promptly reduces cell viability, treatment with K137-E4 has no comparable effect. Only at the highest concentration used (100 μM) does K137-E4 weakly reduce cell viability. Note that this concentration is far above that expected to be effective considering its in vitro IC50 values with CK2. In fact, K137, with a much higher IC50 (Table 1), is already effective in cells at concentrations lower than 10 μM. It therefore seems quite possible that a 100 μM extracellular concentration may permit a minimal penetration inside the cell causing a modest inhibition of the endocellular CK2 (see histogram in Figure 4), resulting, as expected, in slightly accelerated apoptosis.

Viability of Jurkat cells treated with CK2 inhibitors

Figure 8
Viability of Jurkat cells treated with CK2 inhibitors

Cells were incubated for 24 h with increasing concentrations of the indicated inhibitors. Cell viability was assessed using the MTT method, and is reported as the percentage of viable cells compared with the vehicle-treated cells. Results are means±S.E.M. for three independent experiments performed in duplicate. Statistical significance was calculated using an unpaired Student's t test between control and treated cells (***P<0.005; **P<0.01; *P<0.05).

Figure 8
Viability of Jurkat cells treated with CK2 inhibitors

Cells were incubated for 24 h with increasing concentrations of the indicated inhibitors. Cell viability was assessed using the MTT method, and is reported as the percentage of viable cells compared with the vehicle-treated cells. Results are means±S.E.M. for three independent experiments performed in duplicate. Statistical significance was calculated using an unpaired Student's t test between control and treated cells (***P<0.005; **P<0.01; *P<0.05).

Another biological process where CK2 is implicated and which consequently is prevented by typical CK2 inhibitors, is cell motility [5860]. Since vitronectin, one of the known targets of ecto-CK2 [54,55], is a major player in cell adhesion and tissue repair [61], we reasoned that ecto-CK2 activity could be implicated in this process. To assess this possibility, we analysed the effects of the ecto-inhibitor in wound-healing assays. The results, shown in Figure 9, demonstrate that, unlike cell viability, cell migration is inhibited also by K137-E4, supporting a role for ecto-CK2 in this event. The relatedness of the K137E4 effect to ecto-CK2 inhibition is confirmed by failure of AS-E4 and K137-E4 derivatives, devoid of inhibitory efficacy, to affect the wound-healing assay.

Effects of K137-E4 on cell migration

Figure 9
Effects of K137-E4 on cell migration

Cell migration was assessed using a wound-healing assay: HeLa cells were seeded in wells containing an insert, and incubated for 16 h in the presence of increasing concentrations of K137-E4, or AS-E4, K137-E4Me or vehicle, as indicated; then the insert was removed (t=0) and cells were left to migrate for 24 h, always in the presence of the compounds or vehicle (see the Experimental section for details). (A) Representative images from a wound-healing experiment. (B) Quantification of four independent wound-healing experiments performed in duplicate; results are mean±S.D. percentages of wound area covered in 24 h, assigning 100% to the mean value obtained from control cells (***P<0.001, Student's t test).

Figure 9
Effects of K137-E4 on cell migration

Cell migration was assessed using a wound-healing assay: HeLa cells were seeded in wells containing an insert, and incubated for 16 h in the presence of increasing concentrations of K137-E4, or AS-E4, K137-E4Me or vehicle, as indicated; then the insert was removed (t=0) and cells were left to migrate for 24 h, always in the presence of the compounds or vehicle (see the Experimental section for details). (A) Representative images from a wound-healing experiment. (B) Quantification of four independent wound-healing experiments performed in duplicate; results are mean±S.D. percentages of wound area covered in 24 h, assigning 100% to the mean value obtained from control cells (***P<0.001, Student's t test).

DISCUSSION

The results of the present study provide a paradigmatic example of the potential and mode of action of bisubstrate kinase inhibitors, by showing that a moderately potent and quite promiscuous, purely ATP competitive CK2 inhibitor (K137) can be converted into a much more effective and very selective CK2 antagonist by implementing its scaffold with a moiety able to compete for the phosphoacceptor substrate-binding site. Unlike the parent compound K137, which interacts only with the ATP-binding site, its most effective bifunctional derivative (K137-E4) makes multiple interactions with structural elements committed to the binding of both ATP and the peptide substrate. Consequently, its IC50 value for CK2 inhibition is decreased by one order of magnitude and its hit rate, expressing the proportion of kinases that inhibits 50% or more (out of a panel of 140 enzymes) falls from 0.27 to 0.05, a value reflecting an extraordinary selectivity.

Although the experimental approach utilized in the present study is not entirely new and the design of CK2 inhibitors exploiting the same rationale have been already described [18], in at least two respects our data provide novel insights and perspectives. First, the bifunctional mode of binding of our inhibitors not only is supported by the rationale underlying their design, but also has been validated by a mutational analysis unambiguously showing that our compounds interact both with the ATP-binding site and with multiple basic elements responsible for the recognition of the phosphoacceptor substrate, thus perfectly accounting for the mixed-type inhibition kinetics observed. Such a detailed mapping of the interaction network of a bisubstrate inhibitor has never been reported before for other similar compounds, whose bifunctional mode of binding was inferred from structural data where only the interaction with the ATP-binding site was detectable [7,18,19].

Even more noteworthy, from a practical standpoint, is the demonstration that our best bisubstrate inhibitor, K137-E4, thanks to its inability to permeate the cell, has no effect on intracellular CK2 while suppressing the activity of those CK2 molecules which are exposed on the outer surface of the cell. This property makes K137-E4 a valuable tool to shed light on the biological roles of such a pool of ‘ecto-CK2’ and to dissect its functions among those of the population of ecto-kinases present on the outer cell surface.

The existence of ecto-kinase activities is a well-documented phenomenon [62,63], albeit their functions are still rather enigmatic. In the case of CK2, we have shown that its amount is significantly relevant in comparison with total CK2 (Figure 6). The mechanism of CK2 holoenzyme externalization has been investigated in some detail [52,53], and a number of external targets for this subset of ecto-CK2 molecules have been described, including vitronectin [54,55], the C9 complement [56] and collagen XVII receptor [57]. Although the identification of new ecto-CK2 substrates was not among the aims of the present study, we show that several protein bands released into the medium are detectable, which are sensitive to K137-E4 (but not to either AS-E4 or K137-E4Me) and are therefore putative ecto-CK2 targets (Figure 7). Thus K137-E4 can be exploited to dissect the roles and to validate the targets of such a subset of CK2 ecto-kinases without affecting the plethora of CK2 substrates altogether. For the time being, our data show that this pool of external CK2 is not required for the well-recognized anti-apoptotic and pro-survival function of CK2, since cell viability is not significantly reduced by treating cells with K137-E4 under conditions where ecto-CK2 is entirely suppressed (Figures 5 and 8). Rather, our wound-healing assays suggest that external CK2 activity plays a role in cell migration (Figure 9), another process affected by CK2, and may be instrumental to the pro-metastatic potential of CK2 [5860,64]. Although we cannot rule out that the effect of K137-E4 on the wound-healing assay is mediated by a CK2-independent mechanism, this seems extremely unlikely, first due to the very narrow selectivity of this inhibitor, not expected to affect any of the other ecto-kinases present on the cell surface, and, secondly, because the structurally related compound K137-E4Me and the AS-E4 moiety, ineffective on CK2, fail to affect the wound-healing assay as well.

Interestingly the only kinases, besides CK2, whose susceptibility to K137-E4 is increased with respect to the parent compound K137 are those belonging to the PLK family (Figure 2B and Table 4) which have been shown to share with CK2 consensus sequences specified by multiple acidic residues [38,65]. Unlike CK2, however, which is quite sensitive to K137 (IC50 0.25 μM), PLKs are quite refractory to it (IC50 27–40 μM), a circumstance that, on one hand, highlights the dramatic efficacy of the pseudosubstrate derivatization in the case of PLKs (see the K137/K137-E4 IC50 ratio in Table 4), but, on the other, makes PLKs as a whole less susceptible than CK2 to inhibition by the bifunctional ligand (IC50 0.4–0.5 μM compared with 0.025 with CK2). In any case, K137-E4 provides the rationale for the design of dual bisubstrate inhibitors directed towards PLKs, in which the ATP site-directed element, K137, is replaced by the scaffold of one of the ATP-competitive PLK inhibitors already available.

AUTHOR CONTRIBUTION

Giorgio Cozza, Sofia Zanin, Maria Ruzzene and Lorenzo Pinna designed the experi-ments. Giorgio Cozza, Sofia Zanin and Cristina Girardi performed the experiments. Elena Costa, Giovanni Ribaudo and Giuseppe Zagotto synthesized the bisubstrate CK2 inhibitors. Giorgio Cozza performed the in silico analysis. Stefania Sarno and Mauro Salvi provided invaluable reagents. Giorgio Cozza, Maria Ruzzene and Lorenzo Pinna wrote the paper with contributions from all the authors.

We are grateful to Dr Vincent Tagliabracci (University of California, San Diego, San Diego, CA, U.S.A.) for providing Fam20C kinase, to Dr Mario Chiariello (Istituto Toscano Tumori-Core Research Laboratory, Siena, Italy) for providing ERK8, to Dr Izabela Sumara (IGMBC, Illkirch, France) for providing the GST–PLK1 expression vector and to Professor Anna Depaoli-Roach (Indiana University School of Medicine, Indiana, IN, U.S.A.) for providing inhibitor 2 of protein phosphatase 1. We thank The Molecular Modelling Section (University of Padova, Padova, Italy) co-ordinated by Professor Stefano Moro. We gratefully acknowledge the collaboration of Professor Zygmunt Kazimierczuk (Institute of Chemistry, Warsaw Life Sciences University, Warsaw, Poland) who donated K137 and Professor Oriano Marin (CRIBI, University of Padova, Padova, Italy) who provided the peptides.

FUNDING

This work was supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC) [grant number IG 14180 (to L.A.P.)] and by the University of Padova (Progetto Giovani Ricercatori) to G.C.

Abbreviations

     
  • DMEM

    Dulbecco’s modified Eagle’s medium

  •  
  • DYRK1A

    dual-specificity tyrosine-phosphorylation-regulated kinase 1A

  •  
  • ERK8

    extracellular-signal-regulated kinase 8

  •  
  • Fam20C

    family with sequence similarity 20 member C

  •  
  • GSK3β

    glycogen synthase kinase 3β

  •  
  • HEK

    human embryonic kidney

  •  
  • HIPK2

    homeodomain-interacting protein kinase 2

  •  
  • Hsp70

    heat-shock protein 70

  •  
  • PLK

    Polo-like kinase

  •  
  • TBI

    4,5,6,7-tetrabromo-1H-benzimidazole

  •  
  • WB

    Western blotting

  •  
  • WGA

    wheatgerm agglutinin

References

References
1
Lamba
V.
Ghosh
I.
New directions in targeting protein kinases: focusing upon true allosteric and bivalent inhibitors
Curr. Pharm. Des.
2012
, vol. 
18
 (pg. 
2936
-
2945
)
[PubMed]
2
Lavogina
D.
Enkvist
E.
Uri
A.
Bisubstrate inhibitors of protein kinases: from principle to practical applications
ChemMedChem
2010
, vol. 
5
 (pg. 
23
-
34
)
[PubMed]
3
Ekambaram
R.
Enkvist
E.
Vaasa
A.
Kasari
M.
Raidaru
G.
Knapp
S.
Uri
A.
Selective bisubstrate inhibitors with sub-nanomolar affinity for protein kinase Pim-1
ChemMedChem
2013
, vol. 
8
 (pg. 
909
-
913
)
[PubMed]
4
Vaasa
A.
Viil
I.
Enkvist
E.
Viht
K.
Raidaru
G.
Lavogina
D.
Uri
A.
High-affinity bisubstrate probe for fluorescence anisotropy binding/displacement assays with protein kinases PKA and ROCK
Anal. Biochem.
2009
, vol. 
385
 (pg. 
85
-
93
)
[PubMed]
5
Enkvist
E.
Raidaru
G.
Vaasa
A.
Pehk
T.
Lavogina
D.
Uri
A.
Carbocyclic 3′-deoxyadenosine-based highly potent bisubstrate-analog inhibitor of basophilic protein kinases
Bioorg. Med. Chem. Lett.
2007
, vol. 
17
 (pg. 
5336
-
5339
)
[PubMed]
6
Enkvist
E.
Lavogina
D.
Raidaru
G.
Vaasa
A.
Viil
I.
Lust
M.
Viht
K.
Uri
A.
Conjugation of adenosine and hexa-(D-arginine) leads to a nanomolar bisubstrate-analog inhibitor of basophilic protein kinases
J. Med. Chem.
2006
, vol. 
49
 (pg. 
7150
-
7159
)
[PubMed]
7
Lavogina
D.
Lust
M.
Viil
I.
Konig
N.
Raidaru
G.
Rogozina
J.
Enkvist
E.
Uri
A.
Bossemeyer
D.
Structural analysis of ARC-type inhibitor (ARC-1034) binding to protein kinase A catalytic subunit and rational design of bisubstrate analogue inhibitors of basophilic protein kinases
J. Med. Chem.
2009
, vol. 
52
 (pg. 
308
-
321
)
[PubMed]
8
Hines
A.C.
Parang
K.
Kohanski
R.A.
Hubbard
S.R.
Cole
P.A.
Bisubstrate analog probes for the insulin receptor protein tyrosine kinase: molecular yardsticks for analyzing catalytic mechanism and inhibitor design
Bioorg. Chem.
2005
, vol. 
33
 (pg. 
285
-
297
)
[PubMed]
9
Poot
A.J.
van Ameijde
J.
Slijper
M.
van den Berg
A.
Hilhorst
R.
Ruijtenbeek
R.
Rijkers
D.T.
Liskamp
R.M.
Development of selective bisubstrate-based inhibitors against protein kinase C (PKC) isozymes by using dynamic peptide microarrays
ChemBioChem
2009
, vol. 
10
 (pg. 
2042
-
2051
)
[PubMed]
10
Lee
J.H.
Kumar
S.
Lawrence
D.S.
Stepwise combinatorial evolution of Akt bisubstrate inhibitors
ChemBioChem
2008
, vol. 
9
 (pg. 
507
-
509
)
[PubMed]
11
Pinna
L.A.
The raison d’être of constitutively active protein kinases: the lesson of CK2
Acc. Chem. Res.
2003
, vol. 
36
 (pg. 
378
-
384
)
[PubMed]
12
Meggio
F.
Pinna
L.A.
One-thousand-and-one substrates of protein kinase CK2?
FASEB J.
2003
, vol. 
17
 (pg. 
349
-
368
)
[PubMed]
13
Litchfield
D.W.
Protein kinase CK2: structure, regulation and role in cellular decisions of life and death
Biochem. J.
2003
, vol. 
369
 (pg. 
1
-
15
)
[PubMed]
14
Tawfic
S.
Yu
S.
Wang
H.
Faust
R.
Davis
A.
Ahmed
K.
Protein kinase CK2 signal in neoplasia
Histol. Histopathol.
2001
, vol. 
16
 (pg. 
573
-
582
)
[PubMed]
15
Ruzzene
M.
Pinna
L.A.
Addiction to protein kinase CK2: a common denominator of diverse cancer cells?
Biochim. Biophys. Acta
2010
, vol. 
1804
 (pg. 
499
-
504
)
[PubMed]
16
Pierre
F.
Chua
P.C.
O’Brien
S.E.
Siddiqui-Jain
A.
Bourbon
P.
Haddach
M.
Michaux
J.
Nagasawa
J.
Schwaebe
M.K.
Stefan
E.
, et al. 
Pre-clinical characterization of CX-4945, a potent and selective small molecule inhibitor of CK2 for the treatment of cancer
Mol. Cell. Biochem.
2011
, vol. 
356
 (pg. 
37
-
43
)
[PubMed]
17
Battistutta
R.
Mazzorana
M.
Cendron
L.
Bortolato
A.
Sarno
S.
Kazimierczuk
Z.
Zanotti
G.
Moro
S.
Pinna
L.A.
The ATP-binding site of protein kinase CK2 holds a positive electrostatic area and conserved water molecules
ChemBioChem
2007
, vol. 
8
 (pg. 
1804
-
1809
)
[PubMed]
18
Enkvist
E.
Viht
K.
Bischoff
N.
Vahter
J.
Saaver
S.
Raidaru
G.
Issinger
O.G.
Niefind
K.
Uri
A.
A subnanomolar fluorescent probe for protein kinase CK2 interaction studies
Org. Biomol. Chem.
2012
, vol. 
10
 (pg. 
8645
-
8653
)
[PubMed]
19
Pflug
A.
Rogozina
J.
Lavogina
D.
Enkvist
E.
Uri
A.
Engh
R.A.
Bossemeyer
D.
Diversity of bisubstrate binding modes of adenosine analogue–oligoarginine conjugates in protein kinase a and implications for protein substrate interactions
J. Mol. Biol.
2010
, vol. 
403
 (pg. 
66
-
77
)
[PubMed]
20
Schneider
C.C.
Kartarius
S.
Montenarh
M.
Orzeszko
A.
Kazimierczuk
Z.
Modified tetrahalogenated benzimidazoles with CK2 inhibitory activity are active against human prostate cancer cells LNCaP in vitro
Bioorg. Med. Chem.
2012
, vol. 
20
 (pg. 
4390
-
4396
)
[PubMed]
21
Sarno
S.
Mazzorana
M.
Traynor
R.
Ruzzene
M.
Cozza
G.
Pagano
M.A.
Meggio
F.
Zagotto
G.
Battistutta
R.
Pinna
L.A.
Structural features underlying the selectivity of the kinase inhibitors NBC and dNBC: role of a nitro group that discriminates between CK2 and DYRK1A
Cell. Mol. Life Sci.
2012
, vol. 
69
 (pg. 
449
-
460
)
[PubMed]
22
Sarno
S.
Vaglio
P.
Meggio
F.
Issinger
O.G.
Pinna
L.A.
Protein kinase CK2 mutants defective in substrate recognition: purification and kinetic analysis
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
10595
-
10601
)
[PubMed]
23
Sarno
S.
de Moliner
E.
Ruzzene
M.
Pagano
M.A.
Battistutta
R.
Bain
J.
Fabbro
D.
Schoepfer
J.
Elliott
M.
Furet
P.
, et al. 
Biochemical and three-dimensional-structural study of the specific inhibition of protein kinase CK2 by [5-oxo-5,6-dihydroindolo-(1,2-a)quinazolin-7-yl]acetic acid (IQA)
Biochem. J.
2003
, vol. 
374
 (pg. 
639
-
646
)
[PubMed]
24
Venerando
A.
Marin
O.
Cozza
G.
Bustos
V.H.
Sarno
S.
Pinna
L.A.
Isoform specific phosphorylation of p53 by protein kinase CK1
Cell. Mol. Life Sci.
2010
, vol. 
67
 (pg. 
1105
-
1118
)
[PubMed]
25
Salvi
M.
Trashi
E.
Cozza
G.
Franchin
C.
Arrigoni
G.
Pinna
L.A.
Investigation on PLK2 and PLK3 substrate recognition
Biochim. Biophys. Acta
2012
, vol. 
1824
 (pg. 
1366
-
1373
)
[PubMed]
26
Beck
J.
Maerki
S.
Posch
M.
Metzger
T.
Persaud
A.
Scheel
H.
Hofmann
K.
Rotin
D.
Pedrioli
P.
Swedlow
J.R.
, et al. 
Ubiquitylation-dependent localization of PLK1 in mitosis
Nat. Cell Biol.
2013
, vol. 
15
 (pg. 
430
-
439
)
[PubMed]
27
Tagliabracci
V.S.
Engel
J.L.
Wen
J.
Wiley
S.E.
Worby
C.A.
Kinch
L.N.
Xiao
J.
Grishin
N.V.
Dixon
J.E.
Secreted kinase phosphorylates extracellular proteins that regulate biomineralization
Science
2012
, vol. 
336
 (pg. 
1150
-
1153
)
[PubMed]
28
Strambi
A.
Mori
M.
Rossi
M.
Colecchia
D.
Manetti
F.
Carlomagno
F.
Botta
M.
Chiariello
M.
Structure prediction and validation of the ERK8 kinase domain
PLoS One
2013
, vol. 
8
 pg. 
e52011
 
[PubMed]
29
Bain
J.
Plater
L.
Elliott
M.
Shpiro
N.
Hastie
C.J.
McLauchlan
H.
Klevernic
I.
Arthur
J.S.
Alessi
D.R.
Cohen
P.
The selectivity of protein kinase inhibitors: a further update
Biochem. J.
2007
, vol. 
408
 (pg. 
297
-
315
)
[PubMed]
30
Wang
S.S.
p-Alkoxybenzyl alcohol resin and p-alkoxybenzyloxycarbonylhydrazide resin for solid phase synthesis of protected peptide fragments
J. Am. Chem. Soc.
1973
, vol. 
95
 (pg. 
1328
-
1333
)
[PubMed]
31
Berman
H.M.
Westbrook
J.
Feng
Z.
Gilliland
G.
Bhat
T.N.
Weissig
H.
Shindyalov
I.N.
Bourne
P.E.
The Protein Data Bank
Nucleic Acids Res.
2000
, vol. 
28
 (pg. 
235
-
242
)
[PubMed]
32
Friesner
R.A.
Banks
J.L.
Murphy
R.B.
Halgren
T.A.
Klicic
J.J.
Mainz
D.T.
Repasky
M.P.
Knoll
E.H.
Shelley
M.
Perry
J.K.
, et al. 
Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy
J. Med. Chem.
2004
, vol. 
47
 (pg. 
1739
-
1749
)
[PubMed]
33
Halgren
T.A.
Identifying and characterizing binding sites and assessing druggability
J. Chem. Inf. Model.
2009
, vol. 
49
 (pg. 
377
-
389
)
[PubMed]
34
Jacobson
M.P.
Friesner
R.A.
Xiang
Z.
Honig
B.
On the role of the crystal environment in determining protein side-chain conformations
J. Mol. Biol.
2002
, vol. 
320
 (pg. 
597
-
608
)
[PubMed]
35
Cozza
G.
Gianoncelli
A.
Montopoli
M.
Caparrotta
L.
Venerando
A.
Meggio
F.
Pinna
L.A.
Zagotto
G.
Moro
S.
Identification of novel protein kinase CK1 delta (CK1δ) inhibitors through structure-based virtual screening
Bioorg. Med. Chem. Lett.
2008
, vol. 
18
 (pg. 
5672
-
5675
)
[PubMed]
36
Cozza
G.
Girardi
C.
Ranchio
A.
Lolli
G.
Sarno
S.
Orzeszko
A.
Kazimierczuk
Z.
Battistutta
R.
Ruzzene
M.
Pinna
L.A.
Cell-permeable dual inhibitors of protein kinases CK2 and PIM-1: structural features and pharmacological potential
Cell. Mol. Life Sci.
2014
, vol. 
71
 (pg. 
3173
-
3185
)
[PubMed]
37
Cozza
G.
Zanin
S.
Determann
R.
Ruzzene
M.
Kunick
C.
Pinna
L.A.
Synthesis and properties of a selective inhibitor of homeodomain-interacting protein kinase 2 (HIPK2)
PLoS One
2014
, vol. 
9
 pg. 
e89176
 
[PubMed]
38
Salvi
M.
Trashi
E.
Marin
O.
Negro
A.
Sarno
S.
Pinna
L.A.
Superiority of PLK-2 as α-synuclein phosphorylating agent relies on unique specificity determinants
Biochem. Biophys. Res. Commun.
2012
, vol. 
418
 (pg. 
156
-
160
)
[PubMed]
39
Salvi
M.
Trashi
E.
Cozza
G.
Negro
A.
Hanson
P.I.
Pinna
L.A.
Tools to discriminate between targets of CK2 vs PLK2/PLK3 acidophilic kinases
Biotechniques
2012
, vol. 
2012
 
40
Cozza
G.
Bonvini
P.
Zorzi
E.
Poletto
G.
Pagano
M.A.
Sarno
S.
Donella-Deana
A.
Zagotto
G.
Rosolen
A.
Pinna
L.A.
, et al. 
Identification of ellagic acid as potent inhibitor of protein kinase CK2: a successful example of a virtual screening application
J. Med. Chem.
2006
, vol. 
49
 (pg. 
2363
-
2366
)
[PubMed]
41
Copeland
R.A.
Evaluation of Enzyme Inhibitors in Drug Discovery
2005
New York
Wiley
42
Pagano
M.A.
Bain
J.
Kazimierczuk
Z.
Sarno
S.
Ruzzene
M.
Di Maira
G.
Elliott
M.
Orzeszko
A.
Cozza
G.
Meggio
F.
Pinna
L.A.
The selectivity of inhibitors of protein kinase CK2: an update
Biochem. J.
2008
, vol. 
415
 (pg. 
353
-
365
)
[PubMed]
43
Graczyk
P.P.
Gini coefficient: a new way to express selectivity of kinase inhibitors against a family of kinases
J. Med. Chem.
2007
, vol. 
50
 (pg. 
5773
-
5779
)
[PubMed]
44
Di Maira
G.
Brustolon
F.
Bertacchini
J.
Tosoni
K.
Marmiroli
S.
Pinna
L.A.
Ruzzene
M.
Pharmacological inhibition of protein kinase CK2 reverts the multidrug resistance phenotype of a CEM cell line characterized by high CK2 level
Oncogene
2007
, vol. 
26
 (pg. 
6915
-
6926
)
[PubMed]
45
Di Maira
G.
Salvi
M.
Arrigoni
G.
Marin
O.
Sarno
S.
Brustolon
F.
Pinna
L.A.
Ruzzene
M.
Protein kinase CK2 phosphorylates and upregulates Akt/PKB
Cell Death Differ.
2005
, vol. 
12
 (pg. 
668
-
677
)
[PubMed]
46
Kubler
D.
Pyerin
W.
Bill
O.
Hotz
A.
Sonka
J.
Kinzel
V.
Evidence for ecto-protein kinase activity that phosphorylates Kemptide in a cyclic AMP-dependent mode
J. Biol. Chem.
1989
, vol. 
264
 (pg. 
14549
-
14555
)
[PubMed]
47
Ruzzene
M.
Di Maira
G.
Tosoni
K.
Pinna
L.A.
Assessment of CK2 constitutive activity in cancer cells
Methods Enzymol.
2010
, vol. 
484
 (pg. 
495
-
514
)
[PubMed]
48
Ruzzene
M.
Penzo
D.
Pinna
L.A.
Protein kinase CK2 inhibitor 4,5,6,7-tetrabromobenzotriazole (TBB) induces apoptosis and caspase-dependent degradation of haematopoietic lineage cell-specific protein 1 (HS1) in Jurkat cells
Biochem. J.
2002
, vol. 
364
 (pg. 
41
-
47
)
[PubMed]
49
Tagliabracci
V.S.
Pinna
L.A.
Dixon
J.E.
Secreted protein kinases
Trends Biochem. Sci.
2013
, vol. 
38
 (pg. 
121
-
130
)
[PubMed]
50
Tagliabracci
V.S.
Wiley
S.E.
Guo
X.
Kinch
L.N.
Durrant
E.
Wen
J.
Xiao
J.
Cui
J.
Nguyen
K.B.
Engel
J.L.
, et al. 
A single kinase generates the majority of the secreted phosphoproteome
Cell
2015
, vol. 
161
 (pg. 
1619
-
1632
)
[PubMed]
51
Kubler
D.
Pyerin
W.
Burow
E.
Kinzel
V.
Substrate-effected release of surface-located protein kinase from intact cells
Proc. Natl. Acad. Sci. U.S.A.
1983
, vol. 
80
 (pg. 
4021
-
4025
)
[PubMed]
52
Rodriguez
F.
Allende
C.C.
Allende
J.E.
Protein kinase casein kinase 2 holoenzyme produced ectopically in human cells can be exported to the external side of the cellular membrane
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
4718
-
4723
)
[PubMed]
53
Rodriguez
F.A.
Contreras
C.
Bolanos-Garcia
V.
Allende
J.E.
Protein kinase CK2 as an ectokinase: the role of the regulatory CK2beta subunit
Proc. Natl. Acad. Sci. U.S.A.
2008
, vol. 
105
 (pg. 
5693
-
5698
)
[PubMed]
54
Seger
D.
Seger
R.
Shaltiel
S.
The CK2 phosphorylation of vitronectin: promotion of cell adhesion via the αvβ3–phosphatidylinositol 3-kinase pathway
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
16998
-
17006
)
[PubMed]
55
Stepanova
V.
Jerke
U.
Sagach
V.
Lindschau
C.
Dietz
R.
Haller
H.
Dumler
I.
Urokinase-dependent human vascular smooth muscle cell adhesion requires selective vitronectin phosphorylation by ectoprotein kinase CK2
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
10265
-
10272
)
[PubMed]
56
Bohana-Kashtan
O.
Pinna
L.A.
Fishelson
Z.
Extracellular phosphorylation of C9 by protein kinase CK2 regulates complement-mediated lysis
Eur. J. Immunol.
2005
, vol. 
35
 (pg. 
1939
-
1948
)
[PubMed]
57
Zimina
E.P.
Fritsch
A.
Schermer
B.
Bakulina
A.Y.
Bashkurov
M.
Benzing
T.
Bruckner-Tuderman
L.
Extracellular phosphorylation of collagen XVII by ecto-casein kinase 2 inhibits ectodomain shedding
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
22737
-
22746
)
[PubMed]
58
Kim
J.
Hwan Kim
S.
CK2 inhibitor CX-4945 blocks TGF-β1-induced epithelial-to-mesenchymal transition in A549 human lung adenocarcinoma cells
PLoS One
2013
, vol. 
8
 pg. 
e74342
 
[PubMed]
59
Ku
M.J.
Park
J.W.
Ryu
B.J.
Son
Y.J.
Kim
S.H.
Lee
S.Y.
CK2 inhibitor CX4945 induces sequential inactivation of proteins in the signaling pathways related with cell migration and suppresses metastasis of A549 human lung cancer cells
Bioorg. Med. Chem. Lett.
2013
, vol. 
23
 (pg. 
5609
-
5613
)
[PubMed]
60
Zheng
Y.
McFarland
B.C.
Drygin
D.
Yu
H.
Bellis
S.L.
Kim
H.
Bredel
M.
Benveniste
E.N.
Targeting protein kinase CK2 suppresses prosurvival signaling pathways and growth of glioblastoma
Clin. Cancer Res.
2013
, vol. 
19
 (pg. 
6484
-
6494
)
[PubMed]
61
Leavesley
D.I.
Kashyap
A.S.
Croll
T.
Sivaramakrishnan
M.
Shokoohmand
A.
Hollier
B.G.
Upton
Z.
Vitronectin: master controller or micromanager?
IUBMB Life
2013
, vol. 
65
 (pg. 
807
-
818
)
[PubMed]
62
Redegeld
F.A.
Caldwell
C.C.
Sitkovsky
M.V.
Ecto-protein kinases: ecto-domain phosphorylation as a novel target for pharmacological manipulation?
Trends Pharmacol. Sci.
1999
, vol. 
20
 (pg. 
453
-
459
)
[PubMed]
63
Yalak
G.
Vogel
V.
Extracellular phosphorylation and phosphorylated proteins: not just curiosities but physiologically important
Sci. Signal.
2012
, vol. 
5
 pg. 
re7
 
[PubMed]
64
Filhol
O.
Giacosa
S.
Wallez
Y.
Cochet
C.
Protein kinase CK2 in breast cancer: the CK2β regulatory subunit takes center stage in epithelial plasticity
Cell. Mol. Life Sci.
2015
, vol. 
72
 (pg. 
3305
-
3322
)
[PubMed]
65
Johnson
E.F.
Stewart
K.D.
Woods
K.W.
Giranda
V.L.
Luo
Y.
Pharmacological and functional comparison of the polo-like kinase family: insight into inhibitor and substrate specificity
Biochemistry
2007
, vol. 
46
 (pg. 
9551
-
9563
)
[PubMed]

Author notes

1

These authors contributed equally to this work.